108,000 degrees... of phase shift

[earthtonesaudio]

I want to convert an audio signal to an AM signal, phase-delay the HF sine wave, and vary the carrier frequency to get adjustable time delay.  Don't ask me why.

For a 20kHz carrier, a maximum of 15ms delay requires 108000 degrees of phase shift.  All-pass filtering is the logical choice, for its constant amplitude-per-frequency output.

Anyone know of a way to fit 600 all-pass stages into a stompbox?

[gena_p1]

buy panasonic mn3007 on e-bay

[petemoore]

  Why  ?
  ...lol...sorry, had to ask  !

[earthtonesaudio]

Alright, since you asked for it... 

The BBD approach has been done a bazillion times already.  I want to try a different approach, not because I think it would be easier (it won't), and not because I think it will be better (who knows?), but simply because it's different, and I don't think anyone's tried, at least not since BBDs were invented.

Or another, less annoying reason: I think it's possible for this modulation scheme to produce less audible noise than a BBD of comparable delay.

[gena_p1]

600 FETs + 600 Trim Pots + Cooler and you - Columb

[earthtonesaudio]

Perhaps some other filter types could be more efficient with the space...

http://www.filter-solutions.com/description.htg/chevy1bstop.gif
Notice the group delay (pink) varies logarithmically from 3-6MHz (a factor of 2, nice!) while the amplitude is essentially flat.


I wonder if you could achieve a similar response using a long chain of (occasionally buffered) ceramic resonators... Cheap and small...



[edit] Also, the highpass response: http://www.filter-solutions.com/description.htg/chevy2hipas.gif